Core view - focus on major events in the industry at home and abroad, gather the views of Chinese and foreign celebrities and experts, analyze the development of the industry, and take you to understand the future trend!
Jiwei Network reported that in the past year and a half, Nvidia's acquisition of British chip company Arm finally failed. On February 8, the Financial Times quoted three people familiar with the matter as saying that SoftBank's sale of Arm to Nvidia for $66 billion failed after regulators in the United States, the United Kingdom and the European Union expressed serious concerns about its impact on competition in the global semiconductor industry. SoftBank will receive $1.25 billion in compensation fees and seek to push Arm to market in the next fiscal year, the official announcement said.
In fact, as early as late January, there was news that Nvidia was preparing to abandon the acquisition of Arm. Bloomberg reported at the time that the acquisition had made little progress in getting approved due to multiple opposition, and Nvidia had informed its partners that it did not expect the deal to be completed. So, behind the collapse of this largest semiconductor acquisition, what was the original abacus of Nvidia, Arm and SoftBank? Why was this acquisition rounded up by regulators around the world? And what impact will the failure of the acquisition have on the parties?
The abacus behind the acquisition
In terms of the business logic of the acquisition of Arm, Nvidia aims to "drive computing from the cloud, smartphones, PCs, self-driving cars and robots to the edge of the Internet of Things through the combination of its AI computing capabilities and Arm's vast CPU ecosystem, and to promote AI computing to every corner of the world." "Obviously, due to the limited expansion of the GPU computing business and data center market, NVIDIA is committed to expanding the computing power of AI to a wider field and edge to truly lead the era of artificial intelligence."
In addition, whether it is the acquisition of Mellanox or the acquisition of Arm, it can be regarded as an important move for NVIDIA to deal with data center competitors. At present, Intel and AMD have acquired Altera and Xilinx respectively, and have begun to build their own systematic solutions based on "CPU + GPU". At the same time, because the cloud server still has a large number of non-human AI applications running on the CPU, and many AI applications also need CPU support, the acquisition of Arm development server CPU is also a must move for Nvidia.
For Arm, holding a good leg maximizes future opportunities and gets enough technology, resources, financial support, and dynamism and freedom. At present, with the continuous increase of investment, the gap between Arm server chips and X86 server chips is getting closer, and its ecology is also more mature under the market's requirements for reducing energy consumption. By combining with NVIDIA's AI computing capabilities, Arm's business scope will be better expanded into data centers, automotive and network infrastructure.
In addition, because the traditional CPU IP does not support edge AI well, Arm is more and more in the "edge" zone even if it builds a broad ecosystem on the edge side. At present, Arm has developed some IP to support deep learning for AI-based applications, but most companies that develop AI chips are using their own intellectual property to develop DNN modules. Therefore, Arm is not a major player in the market. However, by docking NVIDIA's technology and resources, it will be able to significantly improve its AI capabilities.
In this acquisition, SoftBank's goal is clearly to cash out in a difficult situation. While it has spared no effort in Arm, SoftBank believes the pandemic has further amplified Arm's revenue threat. But even more seriously, SoftBank's Vision Fund had significant financial problems, with multiple investment plans, including WeWork, resulting in an operating loss of about $12.5 billion for SoftBank. As a result, SoftBank sold shares in Alibaba, T-Mobile and other companies to cash out, and went to the node of selling Arm to NVIDIA.
Why is it being encircled and suppressed around the world?
In view of the unique market position of the two sides in the semiconductor field, Nvidia's acquisition of Arm has aroused many oppositions and questions from the beginning. Among the mainstream voices, this will evolve Nvidia into an industry "Big Mac" and will also change Arm's neutrality. Because NVIDIA is an American company, and more than a year of chip shortages have allowed countries to raise the status of semiconductors to unprecedented heights, coupled with the lobbying of competitors, regulators in the United Kingdom, the United States, the European Union and other regulators began to use a "magnifying glass" to look at acquisitions.
In August 2021, the UK Competition and Markets Authority said that the acquisition could disrupt the competitive landscape of the chip industry and consider blocking the acquisition; in October, the EU antitrust regulator issued a warning that the acquisition risked pushing up the global chip price; in November, the UK again announced that it would launch an in-depth investigation into the acquisition, which could take six months; in early December, the European Commission announced that it would suspend the investigation of the acquisition. Meanwhile, the U.S. Federal Trade Commission filed a lawsuit to block the acquisition.
Among these official bodies that announced their investigations, it is not difficult to find some interesting places. First of all, the EU has "gained a reputation" in recent years with a severe anti-monopoly huge fine, but in this acquisition case, it has chosen to "calm things down". Second, the UK's Competition and Markets Authority also said that abandoning Arm would mean losing a powerful "trade and political" weapon. The FTC said Nvidia's acquisition of Arm "will give it the means and incentive to stifle the next generation of chip innovation." It can be seen that all parties have their own considerations.
To a large extent, the fundamental reason for NVIDIA's attempted acquisition of Arm lies in the internal contradictions between large European and American companies. At present, the Arm architecture occupies 95% of the mobile chip market, apple A series, Samsung Orion, Qualcomm Snapdragon, Huawei Kirin, MediaTek Tianji chips, etc. are based on arm architecture. While Nvidia stressed that the acquisition of Arm is aimed at strengthening the data center business, providing users with a "third" option. But it is inevitable that its acquisition of Arm will affect the ecological balance of the entire industry chain.
At present, Nvidia has the world's leading GPU and AI platform, but there is still a gap in CPU and MCU core architecture. These two areas are Arm's stronger areas, and if the acquisition of Arm is completed, it will comprehensively enhance the competitiveness of NVIDIA's products. However, Nvidia itself competes with many of Arm's customers and will be able to contain competitors through the Arm core. This would break the original neutrality, so that it would be difficult to convince all opponents and regulators in any case.
What is the impact of a failed acquisition?
Even if the acquisition of Arm is not successful, it will not have much impact on Nvidia at present. As a well-deserved leading enterprise in the GPU field, NVIDIA also has strong strength in high-performance computing and artificial intelligence. To a certain extent, Nvidia's acquisition of Arm is only a means to enhance its own strength, after all, Arm is mainly involved in the field of mobile chip architecture, and there is a certain difference with NVIDIA's GPU business. In order to strengthen the strength of the CPU, it is rumored that NVIDIA has set up a new CPU research and development team in Israel.
However, in view of the large number of businesses such as data centers, AI, and servers, NVIDIA's lack of Arm may lead to the inability to smoothly achieve business transformation and upgrade expansion, which in turn will lead to a disadvantage in the era of artificial intelligence with Intel and AMD. In the future, the possibility of NVIDIA's transformation to mobile terminal architecture chips is not ruled out, but the failure of the acquisition of Arm will also make the transformation difficult to complete smoothly. In addition, according to the agreement, once the acquisition fails, Nvidia will pay SoftBank a compensation fee of up to $1.25 billion.
For Arm, the failure of NVIDIA's acquisition will affect its own technological innovation and development. As a chip IP manufacturer, Arm needs to maintain a certain expansion ability, while having rapid action and flexibility in innovation. Under the leadership of the capital-deficient SoftBank, Arm may focus too much on short-term revenue and profitability, and cannot focus on promoting the research and development strategy because of the exhaustion of revenue data. Without the support of the rich and technologically powerful Nvidia, Arm will be at a disadvantage in its own innovation and industry competition.
In addition, after arm and X86 architectures, RISC-V, the world's third-largest architecture, has accelerated its rise and created a greater threat. At present, because the RISC-V architecture is more open than the Arm ecosystem, most of the global chip giants use the RISC-V architecture for CPU design and add their own instruction sets to expand. In other words, the relevant enterprises that design chips based on the RISC-V architecture can grasp more initiatives, so the domestic Huawei, ZTE, Alibaba, Loongson Zhongke, etc. have laid out the RISC-V architecture.
While the capital market's interest in high-tech stocks remains, an IPO of Arm is clearly not SoftBank's best bet. On the one hand, SoftBank now wants to be able to cash out its stop loss quickly, and it takes too long to seek to get Arm to go public independently. On the other hand, arm is worth around $20 billion if you follow the usual fast valuation of high-tech companies in the U.S. (8-10 times the revenue of the past year). This is too far from the price code of more than $60 billion given by NVIDIA, which is a huge loss for SoftBank.
epilogue
As a key technology in many fields of the technology industry, the research and development of chips is related to the core scientific and technological competitiveness of a country. Therefore, the major M&A activities in the chip industry are not only related to the future prospects of both parties to the transaction, but also may trigger a chain reaction and weight adjustment of the entire industry, and may even affect a country's right to speak in the global economic system. After triggering the most sensitive audit nerves of regulatory authorities in various countries, it was reported that NVIDIA would abandon the acquisition of Arm, which is actually very in line with the current development trend of events.
But Nvidia also received a huge "windfall.". As the stock price soared, Nvidia's market value exceeded $810 billion at one point in November last year, and although it has been pulled back several times, NVIDIA's current market value is still nearly $620 billion. In any case, with the dust settled on the stumbling acquisition, the chip design industry will remain neutral and competitive, and Arm can continue to provide chip design architecture services to companies. This is undoubtedly conducive to the healthy and orderly development of semiconductors and even the entire technology industry.
However, without NVIDIA, how will Arm cope with the challenges of the era of artificial intelligence? Can we usher in a groundbreaking new situation under the leadership of SoftBank? Without Arm, how would NVIDIA achieve its ambitious ambitions in AI as well as CPU? How to deal with the fierce competition from old rivals such as Intel and AMD? With these questions in mind, we will wait and see in the near future. (Proofreading/Hidden Drei)